The Culmination: MC London

At last, we’ve got the Mythic Championship results. It’s been quite the build up between the ongoing saga of Izzet Phoenix and questions over the proposed mulligan change. We were all looking for definitive answers from London, and while I didn’t get all of them, I think there’s enough evidence to answer some of them. Though not without some caveats.

The Disclaimers

This is the time where I make my usual disclaimer about over-analyzing results from the Pro Tour, now Mythic Championship. This is an invitation-only tournament, so the sample is not random. It is also quite small, so there is opportunity and motive to try and game the field rather than pick on deck based on merit alone. This is also a mixed-format event, and the draft portion has a huge impact on final standings. A player can be mediocre in constructed and flawless in draft for a high finish and vice-versa. Thus, the data should always be viewed skeptically.

However, MC London’s data is particularly disorienting when it comes to assessing the metagame at large. There were a lot of new policies in place that may have skewed the results. The London Mulligan is the most notable, but it’s not alone. Historically, the PT has taken place a few weeks after a new set releases so that players can learn the draft format. But London was a prerelease. Therefore, there was far less time to practice. In turn, this meant draft practice was prioritized over constructed practice, and the Modern results may have been affected as a result.

A Metagame Question

This reprioritization almost certainly exacerbated the problem of pro-level players struggling with Modern. If Twitter is any indication, a lot of players couldn’t find a deck. This isn’t newmost of competitive Magic isn’t Modern so they don’t play that much. Modern is a format that rewards mastery and experience over all else, and if players don’t naturally have that, it is hard to manufacture for one tournament.

For this reason, pros have tended to pick very safe and well-known decks. That London would be Modern was also a relatively late announcement, so it didn’t leave that much time to prepare. In these circumstances, it’s natural for players to just default to whatever decks they’ve had previous success with or appear to be doing well. Rather than try something new, it’s just better to stick with the new hot deck in Izzet Phoenix or last year’s best deck Humans.

Also of note is the London Mulligan’s pre-tournament effects. Before the tournament, all the buzz said that Tron would be the biggest beneficiary, which mathematical models appeared to confirm. Given that Tron is a known good deck and not that hard to play, all the hype and expectation were naturally going to inflate its numbers.

An Information Advantage

The final factor impacting London’s data is open decklists. Wizards likes letting spectators see the decklists and has long been worried about scouting, so they decided to publish all the decklists on Day 1. The only hidden information was how many of each card was in each sideboard. This gave Humans a huge advantage and suppressed combo decks. Humans’ maindeck has been pretty stable for over a year now, so there’s not much difference between hiding and revealing their decklist.

Specifically, the decklist rule makes Meddling Mage much better. Mage is a very powerful card, but only good if you actually name a relevant card. It struggles against decks with multiple key cards or imperfect information. Forest, go could mean a slow start from Elves, Hardened Scales, and Tron or Amulet Titan. A turn 2 Mage in the dark is a wild guess. With access to opponents’ decklists, each blind Mage was at least in the ballpark, and far more potent.

Additionally, combo decks were denied the surprise boost from opponents’ confusion. There are a lot of very powerful combo options in Modern, and at least some players may have been trying to break the format with the London Mulligan. However, the threat of Humans combined with their deck’s reliance on surprise was killed by open decklists. The incentive was to fall back on safe, known lists, rather than branch out.

Day One

So when it comes to actually looking at the data from London, there really isn’t much surprise. According to Wizards, Tron made up 14.6% of the field, followed by Izzet Phoenix at 12.0% and Humans with 10.3%. There is a steep drop-off to fourth place with UW Control’s 7.4%. The expectation that the more well-known and popular decks would do well is holding strong. Humans was last year’s big deck, and Izzet has been this year. What I didn’t expect was Tron to show up as strongly as it actually did. Beating out Izzet Phoenix for most popular deck is no small feat, and I suspect happened on the strength of the hype around Tron benefiting from the new mulligan. Other than that, the metagame looks quite diverse for an invitational tournament.

Day Two

The Day 2 data is rather unexceptional, as every deck made Day 2 in numbers equivalent to its initial population. The win rates are all over the place, but their order in terms of total representatives didn’t change at all. This doesn’t indicate any noticeable change in strength. Focusing on conversion rates is something of a fallacy, since popular decks will suffer from their popularity. This is borne out in the data, since the lower the starting population, the higher the conversion rates.

In Context

When it comes to overall win percentages, things get more interesting. Despite expectations set by Humans’ advantage, the best performer was Ad Nauseam. Tron is indeed a great matchup, but Humans is quite bad. I presume that expert sideboarding is the reason for this jump. Overall, no deck really shone. The average win rate is roughly 50%, so Tron’s 47.7% is not very good. This is likely due to players expecting the deck and packing enough Damping Spheres to protect themselves.

The Top 8

Humans won, which is hardly surprising considering there were three copies in the Top 8. Tron managed two, with TitanShift, Affinity, and Izzet Phoenix rounding it out. Tron and Human’s numbers are hardly surprising considering their Day 2 presence, nor is Phoenix placing. It is proving very hard to keep the firebirds down. It’s impressive that Matt Sperling made it with Affinity since it’s been displaced by Hardened Scales. I imagine the Experimental Frenzys were critical to his success.

Given that, the interesting part of the Top 8 is that no UW Control made it. UW has a close matchup against Izzet Phoenix, can be made to handily defeat Tron, and fairly good against the field. I imagine it fell foul of Humans. UW doesn’t play that many unique creature removal spells so it’s easy to Meddling Mage it out of the game. You can also prevent a miracled Terminus by Vialing in Kitesail Freebooter in response to the trigger. Jeskai had a far better matchup, but it’s weaker against the field.

Despite all this, I wouldn’t read too much into the Top 8. Given that draft results are included, the actual finishes aren’t very indicative of a deck’s strength. To look

The Modern Story

Instead of the final placings, I would look at just the Modern match points. Doing so eliminates any boost or drag from the draft, and gives a clearer picture of deck strength. No deck ran better than 9-1 in Modern. This is the metagame of 8-2 or better decks.

Deck NameTotal #
Izzet Phoenix5
Humans4
Dredge2
Ad Nauseam2
Red Eldrazi2
UW Control2
Tron2
Burn1
Jund1
Titanshift1
Hardened Scales1
Affinity1
Amulet Titan1
Whir Prison1

This table looks very similar to the metagame we’ve been observing for the past several months. Izzet Phoenix is clearly on top, with Humans nipping at its heels. The rest of the metagame is very broad and roughly equal in metagame share, and therefore power level. To keep beating the dead horse, this indicates that the overall metagame is very healthy. The question remains: is Phoenix’s continued seat atop the standings still acceptable? I think the answer has become no.

Despite its numbers being relatively down from GP levels, Phoenix is maintaining a high win rate. The second most popular deck, it had a 52.7% win rate. That’s pretty average for the field. However, popular decks should necessarily have a lower rate, because they have more mirror matches and pilots that bomb out, cancelling some wins of the successful. Ad Nauseam did the best with 61.7%, and only boasted eight copies in the tournament. So only one player had to do well for the rate to bloom.

However, despite appearances, Izzet Phoenix’s win rate is staggering considering that players were running a ridiculous amount of Surgical Extractions in London. The card that players are convinced is critical to beating Phoenix, despite everything, was the most popular card in London, and five Phoenix players still managed 8-2 or better. Granted, there are other (better) options for defeating Phoenix, but Surgical’s speed and splashability, as well as the perception players have of it being great against Phoenix, continue to make it a favored sideboard choice. In terms of other hate, Relic of Progenitus was 15th place and Path to Exile 4th. Players were clearly aware of Phoenix, prepared for it, and yet Phoenix was still winning more than everyone else. Given that the field was extremely hostile, a 52.7% is very high. This makes me think Phoenix is running out of time.

My Take

I predicted that the metagame was relatively settled going into London, with Phoenix on Top followed by Humans with the rest of the meta trailing in their wake. I didn’t account for Tron’s popularity, but other than that London’s results do look very similar to GP Yokohama‘s which were consistent with all the earlier results. Thus it’s safe to say that Izzet Phoenix is clearly on top of the metagame with Humans, Tron, and Grixis Death’s Shadow being strong contenders.

As for Phoenix itself, I’ve remained hopeful that the metagame would naturally adjust and push it back down. That doesn’t appear to be happening. Izzet Phoenix continues to show up in large numbers and then turn those into high tournament placings. Couple that with a high win rate in a very hostile field, and it’s hard to argue that Izzet Phoenix can’t overcome any obstacle. This moves the deck from an interesting anomaly into a metagame trend, and potentially a dangerous one.

I don’t think it’s inherently broken by any means. However, all the data is indicating that it’s taking up a worrying amount of Day 2 space and is arguably dominating Modern. This often warrants action. It could be winning too early too often as well, but only Wizards knows if that rumor is true. I imagine based on available data that action will be taken against Izzet Phoenix. My guess is a Faithless Looting ban to nuke the best starts.

The Mulligan Question

This begs the question of whether the London Mulligan will be sticking around. The simple answer is that I don’t think the Mythic was a very good test. As a result, I can’t determine anything conclusive about the mulligan change. The fundamental problem was that there were too many variables at play per my disclaimers. If this weren’t a prerelease draft, pros would have had more time to test Modern. This could have led them to make less safe deck choices.

Open decklists also punished players looking to branch out. A lot of the value of picking a wonky rogue deck is opponents being confused by something unexpected. That wasn’t possible. Therefore, London was not really a test of what the mulligan was capable of because there was no incentive to push the envelope. As a result the rule didn’t get a very rigorous field trial, just a nice safe unveiling. Thus I don’t consider it valid data.

The real test is ongoing (at time of writing) via MTGO. Unless Wizards has a massive change of heart, we’ll never know the exact effect of the mulligan because we’ll never see the data. If testimonials saying the change is fine are correct, the rule will stick. If Ken’s is more accurate, then it may not. Many definitely think the rule is good, and apparently Mark Rosewater said in a panel that he expected it to stick. Therefore, I expect that it would take a major distortion in the MTGO metagame to prevent this change.

What Comes Next?

Even if Wizards doesn’t change the mulligan rules nor ban anything from Izzet Phoenix, there are big changes on the horizon for Modern. Specifically, Modern Horizons’ spoiler season begins later this month. Even if the set’s impact is muted, as I expect, it will still shake things up and is guaranteed to bring in some fresh blood. I look forward to seeing what shakes out soon.

Leave a Reply